Contingency management interventions are quite successful at helping people change problematic behavior and in adopting behaviors related to the development of healthy lifestyles. One area in particular in which contingency management interventions have become a frontline treatment is for substance abuse. Contingency management interventions have been repeatedly demonstrated to be successful at initiating periods of abstinence from many different drugs of abuse in a variety of clinical samples. Typically, contingency management interventions .for the treatment of drug abuse involve providing positive reinforcement for the provision of biological samples (e.g., urine or breath) that indicate recent abstinence. Generally these reinforcers are delivered each time a negative-sample is provided. However, basic science research into the delivery of reinforcement suggests that reinforcing behavior less frequently may, in fact, produce behavior that is more persistent once reinforcement has been discontinued. Similarly, basic research suggests that reducing the predictability of reinforcement promotes persistence of behavior change once reinforcement is withdrawn. These observations suggest a strategy for producing long-term behavior change. In substance abuse, this long-term behavior change could be reflected as long-term abstinence. The study outlined in this application is designed to examine whether less frequent and less predictable reinforcement of methamphetamine-negative urines leads to longer-term behavior change than more frequent and more predictable reinforcement. To accomplish this, we propose to compare 4 groups, all of which will receive a standardized psychosocial substance abuse treatment: (1) standard psychosocial treatment, (2) continuous contingency management: reinforcers available after each provision of a methamphetamine-negative urine test, (3) intermittent predictable contingency management: reinforcers available after the provision of every three consecutive methamphetamine-negative urine samples, and (4) intermittent unpredictable contingency management: reinforcers available on one day each week (randomly selected after the first week) for the prevision of methamphetamine-negative urine samples provided all urine samples since delivery of the last reinforcer were also methamphetamine negative. Based on the widespread success of contingency management across many different applications, we believe that the results of the proposed study will have 9enerality beyond the field of substance abuse treatment.